Implementing New Practice: The Roles of Translation, Progression and Reflection

ABSTRACT The aim of this article is to assess the contribution provided by the instrumental theory of translation when conceptualizing micro processes of change. A core assumption of this framework is that the translation of management ideas is characterized by regularities, or translation rules, and that performance of translation rules depends on translation competence. The assessment is based on an exploration of how the process of providing contents to a new model of manpower planning among doctors in a Norwegian health trust unfolded. Participant observation and documents represent the main sources of data. The analysis demonstrates that project methodology competence emerges as a key factor of influence on the performance of translation rules, leading to the interpretation that there is a ‘taken-for-grantedness’ in change implementation which suggests additional competences at play. We propose that the instrumental theory of translation would benefit from extending the framework for contextualizing the performance of translation rules.


Introduction
The view of planned change in organizations has moved along a continuum from Lewin's (1951) model of linear phases, anticipating a beginning and an end for the change at stake, to a more processual way of understanding change, considering change to be a continuous and fluid process (Hernes et al., 2015;Weick & Quinn, 1999). In line with this development, the importance of directing attention to the micro-processes of change is emphasized (Balogun & Johnson, 2005;Gondo & Amis, 2013;Greenwood & Hinings, 1996;Guiette & Vandenbempt, 2017;Liguori, 2012;Maholtra & Hinings, 2015;Tsoukas & Chia, 2002), for example when addressing recurrent patterns as the low success rate of organizational change initiatives (Buchanan et al., 2005;Burnes & Jackson, 2011;Smith, 2002), and the tendency for intended strategies to lead to unintended consequences (Balogun & Johnson, 2005).
In this article, we elaborate further on how to understand micro processes of change by assessing the instrumental theory of translation (Røvik, 2016). This theory is a new contribution to the 'translation perspective' (Waeraas & Nielsen, 2016), in which the fundamental role of micro processes of social interaction (Callon, 1986;Latour, 1986) represents a main point of departure. The instrumental theory of translation is located within the framework of Scandinavian institutionalism (Czarniawska & Joerges, 1996;Czarniawska & Sevón, 1996Sahlin-Andersson & Engwall, 2002). Here the inherent premise is that change is closely associated with the translation of management ideas and models (Sahlin & Wedlin, 2008;Waeraas & Nielsen, 2016), indicating that the contents of change and how it is approached is the main subject of interest. According to Gondo and Amis (2013) translation studies (Boxenbaum, 2006;Lamb & Currie, 2012;Lamertz & Heugens, 2009;Pipan & Czarniawska, 2010) have shed light on implementation, and as such energized the field of implementation studies. A main focus in this approach has been on variations in how new versions of organizational ideas are translated in the local context (Gondo & Amis, 2013).
However, a number of contributions also direct attention to how translations can be characterized through regularities described as a set of editing-or translation rules (Røvik, 2007(Røvik, , 2016Sahlin & Wedlin, 2008;Sahlin-Andersson, 1996). According to Sahlin and Wedlin (2008), the regularities stem from the institutional setting in which the translations are performed. Thus the translations reveal rules that have been followed. Røvik (2016) takes the assumption further by presenting a new instrumental theory of translation, where the ambition is to bring the translation work characterized through regularities even closer to the fore. Aiming to assess the theoretical framework, the research question asked in this article is as follows: How does the instrumental translation framework account for micro-processes of planned change?
In addition to presenting his own set of translation rules that display themselves in translation processes, Røvik (2016) introduces the assumption that a presence of 'translation competence' provides the opportunity for the translators to make informed choices regarding translation rules. The combination of the concepts of translation rules and translation competence, creates a space for discussing whether the use of certain translation rules is informed by knowledge concerning the contents of the idea, i.e. translation competence. Through this theoretical elaboration, the aim is to show why informed choices are important and should take place.
In spite of the considerable attention paid to translation in the research literature on change, more in-depth understanding of translation work is still called for (Cassell & Lee, 2017;Morris & Lancaster, 2006;Mueller & Whittle, 2011). For example, Mueller and Whittle (2011) state that 'What is missing from the literature is a sophisticated set of concepts to understand the detailed, micro-level interactions through which ideas are translated ' (p. 189). This message is repeated in the article of Cassell and Lee from 2017. There have also been specific calls for more research on what occurs when translation rules are performed (see, for example, Teulier & Rouleau, 2013). Waeraas and Sataøen (2014) point to a need for further empirical studies that 'focus on the regularities of translations, the variations in rules, and how their outcomes are connected with 'real time' translation work ' (p. 251). The concept of translation rules has also been applied in several studies (Lamb & Currie, 2012;Magnussen & Tingvold, 2015;Nielsen et al., 2014;Waeraas & Sataøen, 2014). However, to our knowledge, the deployment of Røvik's (2016) framework that relies on both translation rules and translation competence has been scarce in empirical analysis. For instance, in a recent published article (Øygarden & Mikkelsen, 2020), the authors refer to Røvik (2016) both when using the concept of translation rules and when they articulate the need for research on what facilitates good translations in order to attain organizational goals. Interestingly, however, they do not turn to translation competence to find the answers. Instead, they combine the concept of readiness for change (Armenakis & Harris, 2002) with editing practices (Teulier & Rouleau, 2013) in order to get access to translation performance.
On this background, this study sets out to assess the instrumental theory of translation through an empirical investigation. We study a working group that was established and assigned the task of providing contents to a new model of manpower planning among doctors in a Norwegian health trust. This is the initiating stage of an organizational change, which is a crucial step in translation (Czarniawska, 2009). Our discussion is based on how the process of providing contents to a new model of manpower planning unfolds. A particular interest then, is to identify the translation rules performed during the process, and assess how the performance is related to translation competence. As such, we seek to take the call for more research on the performance of translation rules further (Cassell & Lee, 2017;Waeraas & Sataøen, 2014).
The study contributes on a theoretical and a practical level. At the theoretical level, some critical remarks and further developments of the instrumental translation theory are suggested. A point of importance is that additional competences should be accounted for when addressing micro-processes of change within the frames of the instrumental theory of translation. At the practical level, the study directs attention to a 'taken-for-grantedness' in change implementation which might impact how new ideas are being translated. A suggestion is for change agents to raise the awareness of both conducting translations, and of competences in use.
The article is structured as follows. First is a description of the idea under translation. Second, the analytical framework of the study is outlined. Third, we explain our methods of data collection and analysis. Fourth, our findings are presented. Finally, we discuss the results and present our contributions.
The Idea Under Translation: Activity-based Manpower Planning (AMPP) Manpower planning in hospitals can hardly be considered neither a new nor a particularly popular idea travelling around in the organizational field of specialist health services at the moment. Rather, manpower planning is part of a well-developed system producing work-and shift schedules based on legal frameworks and systems of agreements (Øhrn & Torvbråten, 2015). However, there is currently a shift in how this planning is approached (Birch et al., 2020), appearing in the spread of a new best practice (Szulanski & Jensen, 2006) or management idea (Morris & Lancaster, 2006).
In our health trust the new model to be implemented was called Activity-based ManPower Planning (AMPP), and was previously implemented in another health trust in Norway. An interdisciplinary working group was put together, mainly consisting of doctors and administrative staff. The group was assigned to concretize the contents and practice of AMPP in the health trust. Based on their work and recommendations, implementation of the model would take place throughout the hospital departments. The following quote illustrates how the expectations to the new model were described by the top management and the board during the process of considering AMPP: A shift towards activity based manpower planning means that the work plans must be based on knowledge of expected activity together with available resources and limitations in a given period of time. Available resources and limitations refer to budget, available manning within the different professions, and planned absence (vacation, courses, leave of absence). This shift implies that the work schedules must be based on another methodology than the one used today (present practice is a static and a rotating work schedule covering a certain amount of weeks and is repeated an unlimited number of timesin many occasions for several years). (Excerpt derived from case for the board of the health trust).
Another quote elaborates in more detail on the contents of the new model and shortcomings in current practice: AMPP means planning long term regarding activities and manning in addition to a limitation of individual choices associated with work planning. This has previously been delegated to the doctors themselves, and in some occasions to employee representatives. A lack of experience among leaders regarding doctors' work planning has been revealed, and also that such efforts among leaders have been regarded as a challenge to the doctors' traditional autonomy. The CEO is still of the opinion that a better control and coordination of the doctors' work plans will have considerable economic impact. (Excerpt derived from case for the board of the health trust).
According to this new idea, it is essential to base the planning on expected activities, and not only on resources. In principle, expected activity in terms of contents and duration could be estimated according to both the individual doctor's experience and tacit knowledge and the available historical data. Based on the intention of more control at the expense of individual autonomy, however, the version of AMPP described above includes an expectation of more standardized tools to measure and plan activity, implying the use of statistics and numbers. It is also essential that the plans must provide an accurate picture of the presence of the doctors, i.e. take into account scheduled absence such as holidays, courses, etc.
When we later on in the analysis refer to the original version of AMPP and the intentions with it, it is the description of contents above we have in mind. For our purpose, the original version of AMPP equals existing problem definitions by the top-management team, and how AMPP is formulated as a written description of the idea in several cases for the board leading to decision of implementation.

Theoretical Framework
Our investigation of how the change process unfolded, starts out by assuming that the concept of translation rules is essential for describing and understanding the development of contents in change processes. We outline this assumption and how the performance of translation rules is linked to translation competence, starting with an introduction of the premises of the instrumental theory of translation.

The Instrumental Turn and Translation Competence in Translation Theory
The instrumental theory of translation (Røvik, 2016) is an effort to develop further the potential of translation theory when analysing transformation of management ideas. In order to do this, Røvik turns to translation studies, which connects several disciplines, for example, linguistics, literary theory and communication theory (Snell-Hornby, 2006). Within this body of research, the assumption of regularities in translation represents one of three features that, according to Røvik (2016), characterize the research tradition and make it interesting as a source of inspiration. Deploying the assumption of regularities on the translation of managerial ideas, his argument is that such translations may display regularities that can be analysed and subsumed under more general translation rules, just like translations of cultural artefacts, including language, images, and symbols, adhere to basic patterns (Røvik, 2007(Røvik, , 2016. The attention paid to 'good translations' and communication between a source and a target are presented as additional features of inspiration received from translation studies. By referring to the focus on 'good translations', Røvik (2016) emphasizes the need to pay more attention to the probable outcomes of translation processes, and also how to increase the probability of achieving organizational ends. Through this move, the framework enters its instrumental direction, and can be argued to relate to a search for predictability that is part of the vast majority of social science (May & Finch, 2009). This rational object of science has however been actively rejected in the methodological underpinnings of translation theory (Law, 2004), indicating that the framework of the instrumental translation theory moves away from fundamental premises of translation theory as it is reflected in Scandinavian institutionalism.
Through the reliance on communication between a source and a target, close connection to the knowledge transfer literature (Gupta & Govindarajan, 2000;Szulanski et al., 2016) is further revealed in the development of the framework. The elaboration of the translation competence concept, for example, appears as heavily reliant on knowledge management and knowledge transfer concepts, such as tacit and explicit knowledge, features of the transferred knowledge, and features of the relation between recipient and source (Szulanski et al., 2016). On the whole then, and instead of building on existing and re-known contributions to neo-institutional theory, important building blocks are taken from other research traditions. A question that can be raised, but which is not addressed further here, is whether we observe an imbalance in Røvik's attempt to on one hand setting out to investigate further the potential of translation theory, on the other, locating the investigation of the new framework in the border area of Scandinavian institutionalism.
Turning back to the attention paid to regularities of translation in this article, and which is the starting point of our empirical analysis, further details about translation rules need to be introduced. The term 'translation rule' points to more or less explicit guidelines for the appropriate translation of knowledge (Røvik, 2016, p. 7). Four translation rules that display themselves in the translation of managerial ideas are suggested: copying, addition, omission, and alteration. Application of the rules takes place within the frame of a particular mode: the reproducing (copying), the modifying (addition and omission) or the radical mode (alteration). Each of the modes is referring to a distinguishable intention and a style of translation performance, and presupposes that the translators of a certain change initiativewhen using translation competencechoose one specific mode, and stick to it during the whole process. However, the modes are sometimes taken for granted, echoing habits, culture and traditions in the translator's context.
Given the rule-based characteristics of translation processes, a possible instrumental turn in translation theory is made possible (Røvik, 2016). Such a turn draws the attention to the translators using their translation competence in order to translate practices and ideas between organizational contexts in ways that increase the probability of achieving organizational ends. Røvik (2007Røvik ( , 2016 states that translation competence is becoming more and more important as a strategic organizational resource. This is due to the increasing production and spread of organizational ideas, along with an expectation of transferring best practices between organizations (Benders et al., 2007). Accordingly, deliberate use of translation rules, based on translation competence, represents a core dimension of successful change implementation.
Translation competence has as its core knowledge about the idea being 'exposed' to translation. More specific, translation competence has three components: knowledge of the idea to be implemented, knowledge of the local practice where the new idea is being translated, and knowledge of the translation rules (Røvik, 2016). For example, knowledge of the idea may imply knowledge of the complexity of the idea, the idea's degree of tacit and/or explicit components, etc. (Lillrank, 1995;Røvik, 2016;Szulanski, 1996). Concerning knowledge of the new idea, the assumption in our case is that members of the working group will either have such knowledge of AMPP in advance or acquire it. Knowledge of the local practice implies knowing the 'old' practice of manpower planning that is going to be replaced, and the local context of the doctors' job situation. In order to describe these components of translation competence further, Røvik introduces the notion of translation rules' scope conditions, i.e. the conditions under which it will be reasonable to apply which rule (Røvik, 2016, p. 11). The components are applied together to choose the translation rule(s) most suited to achieve organizational ends.
Our ambition is to assess whether the possible performance of translation rules can be ascribed some sort of orientation towards the three above listed components. This is a first step in trying to make use of the whole framework, and based on our analysis make some reflections of its possible fruitfulness. In our empirical analysis, however, we do not set out to 'map' competences present in the group, neither do we relate to the debate on whether competence is something that individuals have or is created in practice (Brown & Duguid, 2001). We limit ourselves to investigate how the process that is studied displays the performance of translation rules, and as such calls for a reflection on the presence or lack of translation competence.
While demonstrating the performance of translation rules, a main observation in this analysis is that matters related to the change design appeared influential. The observation 312 E. A. NILSEN AND A. G. SANDAUNET implies that the field of implementation studies is brought into focus (Al-Haddad & Kotnour, 2015;Balogun et al., 2015;Beer & Nohria, 2000;Kotter, 1995;Meyer & Stensaker, 2006;Stace & Dunphy, 1991), preparing the ground for considering additional competences when contextualizing translation rules.

Change Design as an Additional Area of Competence
Change designs are normally placed along the axis of being top-down or bottom-up oriented, and they are assumed to impact implementation processes with basis in their respective orientation (Al-Haddad & Kotnour, 2015;Beer & Nohria, 2000). Within translation studies, however, the impact of change design has not been addressed. In this study, we approach and describe components of change designs, extending the potential competences at play. Theory E is presented as a version of a top-down approach and theory O as a version of a bottom-up approach (Beer & Nohria, 2000). In his article, Leppitt (2006) draws parallels between theory E and different approaches to project management that have emerged from engineering industries. He claims that this focus on project methodology to drive change is a useful reference point for theory E change models, or «hard» approaches to change management. This is supported by other studies on project organizing (Fred, 2019). In this study we suggest project methodology to be a competence area of relevance when discussing the performance of translation rules. Projects are created through delimiting a specific task, goal, time, and team (Lundin & Söderholm, 1995). Temporariness functions as an important premise provider for how work is planned and coordinated within projects (Nordqvist et al., 2004), in the sense that working towards deadlines and within a predefined timeframe, force to the fore such factors as a clear mandate, clear goals, requested output, reporting structures involving a steering committee (being the link to top management), and a project manager (Kerzner, 2013). All the elements are put together in order to make sure that progress towards the final result is taking place. Following this line of argument, project methodology competence can be argued to display itself through emphasizing project progression.
Theory O on the other hand, has a somewhat different approach (Beer & Nohria, 2000), and directs attention to a third area of competence. This theory emphasizes that the main goal is development of the organizational capacity or learning ability. In order to succeed, one has to facilitate processes that allow the participants to be involved (Eriksson et al., 2016;Hodson, 1996;Weick, 2000). Involvement appears in various forms depending on context. For our purpose, we find support in contributions which focus on the concept of public reflection as a premise for organizational learning (Raelin, 2001). According to Raelin (2001, p. 11), reflection is 'the practice of periodically stepping back to ponder the meaning of what has recently transpired to us and to others in our immediate environment'. Raelin further considers public reflection as the key to unlocking the learning from project-based work and describes it as associated with learning dialogues. The value of stepping back is also emphasized by Malhotra and Hinings in their study of change in three case firms (2015). Related to our context we ask whether reflection emerges as yet another area of competence that must be considered.
Our theoretical framework is summarized in Table 1.

Method and Material
Our investigation is conducted as trailing research (Finne et al., 1995;Olsen & Lindøe, 2004). Trailing research combines active participation in some phases of an intervention with tools as passive observation, standardized methods and document analysis (Olsen & Lindøe, 2004). In this research, we participated in group meetings, mainly as passive observators, but occasionally also through more active contributions. In a wider methodological context, our approach is rooted in the broad category of social constructivist research in which a view on the researcher as a co-producer of data is integrated in the research frame (Alvesson & Sköldberg, 2000). The context bounded character of such data and the surrounding debates on utilization and generalization of the data (Sandelowski, 2004), is not addressed in this article. Our aim here is to use the insight provided by the study to raise further issues on change management into debate. However, questions related to the quality of the data (Tracy, 2010) are still raised and commented on in the following outline of how this research was conducted.

The Research Context
Eight persons participated in the working group. There were two administrative representatives from the central HR unit, one administrative representative from an HR-section working with statistics and analysis, one administrative consultant working in a department where changes in the doctor manpower plan had been addressed as part of this department's quality improvement work, two doctor employee representatives and two doctors from selected departments. One of the representatives from the central HR unit entered the position as group coordinator, and also reported to a steering committee. Ethical approval was provided by the Norwegian Centre for Research Data.

Material
We collected our material through participant observation and document studies. The working group held sixteen meetings between September 2014 and January 2015.
Both researchers attended twelve meetings, while only one researcher was present at three meetings. At one meeting (number five) none of us could attend. The steering committee held three meetings during the project period. Both researchers attended to one of the meetings, while one of us attended to the two other meetings in the committee. The discussions in the working group represent our main source of data. We asked for permission to record the group meetings during our initial contact with the group coordinator, but active consent was not provided by the group participants. As such, recorded material from group discussions is not included. The observational data are thereby based on the written notes of each researcher, and represent the 'transcription' we produced from the meetings. We endeavored to provide a transcription of the conversations that was as detailed and accurate as possible, for example by constantly working to keep track with who was saying what, and to repeat the contents in full sentences. Additional material was included as observational data. We had an ongoing dialogue, or field conversations, with the group coordinator. These conversations were mainly conducted in terms of informal talks after group meetings. However, one of them was scheduled and initiated by the group coordinator. The initiative was a response to the preliminary thoughts we had offered on the process in a presentation held at meeting number seven. Notes from our talks with the group coordinator supplemented our observational data. In addition, e-mail communication between the group coordinator and the steering committee was made available for us and is included as observational data derived from the group process.
Documents that are included in this analysis range from a description of the mandate of the group, minutes of meetings that were sent by the group coordinator, power point presentations that were held by the group coordinator, presentations conducted by group participants and the report that was delivered by the group at the end of the process.

Analysis
Our analysis of the empirical material rests on the framework of directed content analysis (Hsieh & Shannon, 2005). Directed content analysis is deployed when the aim is to validate or extend conceptually a theoretical framework or a theory, such as this assessment of the instrumental theory of translation. A chronological reconstruction of the group process that included the notes of both researchers represented the initial step of this analysis. It was followed by a thematic coding (Gibbs, 2007) of the contents of the group discussions. Seven themes, or categories, both data and researcher generated (Gibbs, 2007), were identified: 'AMPP', 'current practice of manpower planning', 'other processes in the trust', 'group coordination', 'the rostering system', 'use of statistics', and 'group processes'. A table was then elaborated in which the meetings were set up in a chronological order along the horizontal axis, and the seven themes along the vertical axis. Based on this division, we could write keywords for the contents of each meeting. The approach implied that some meetings were filled with keywords on two or three themes, while others had more themes addressed. A number of topics could be discussed within the frames of each theme.
A focus on 'sequences of talk' (Heritage, 2004;Stivers & Sidnell, 2013) was then included and directed attention to the topics that were introduced and discussed. The concept sequences of talk is borrowed from the framework of conversation analysis (CA). Sequences of talk are constituted by the topics that are introduced in a conversation, and the responses given to these topics. Each sequence is 'opened' and 'closed', which makes it separable from other sequences. Between the opening and closing, an idea or suggestion receives responses that can be supportive or critical, or even ignorant (Heritage, 2004). In the context of this analysis, the response provided to the topics that were introduced is the entrance to identifying the translation rules that received support in the working group. One example is when the potential of using statistics on activity, or historical data, in estimating future activity, is introduced as a topic for discussion. Expressing support to suggested potential of using statistics, is considered as a response that contributes to a copying of what is here described as the original idea of manpower planning. An utterance of skepticism, on the other hand, will be considered as a response that contributes to a rejection of the original idea. The sequence 'closes' when the attention turns to another topic, for example without addressing further the skepticism uttered to a suggestion raised by a group member. Our focus on sequences also implies that attention is paid to which topics that are introduced and responded to in group discussions. An important finding in our analysis is that ignorance, or not 'picking up' acknowledged topics for further discussions, became a notable part of the group process.
The attention paid to sequences of talk in this analysis relates to a category of analytical methods that provide insights into micro-level of interactions in everyday life and institutional contexts (Heritage, 2004). Mueller and Whittle (2011), for example, deploy 'discursive devices analysis' in order to analyse translation work in an organizational context. Compared to this article's focus on openings, responses and closing of different topics, discursive devices analysis is more directed towards the 'sender' or 'initiator' of a topic for example, the change agentand the linguistic devices used by this sender. Our focus on sequences of talk, on the other hand, directs more attention towards the response provided to the sender, and thereby the translations that gain support and are performed through these interactions.
While being aware that analysis of sequences of talk in the CA-tradition is strictly connected to recorded material (Heritage, 2004;Stivers & Sidnell, 2013), we used hand written notes as the point of departure for identifying sequences with openings and closings. Lack of recorded material reduces the level of detail in transcription and thereby the transparency of the research (Tracy, 2010). As a consequence, the basis for interpretations becomes less solid and illustrates limitations of this research.
Our interpretations still rely on different sources of data, for example, illustrated by the e-mail communication between the steering committee and the group coordinator that was made available to us. As such, we were provided access to additional material that contributed to reveal the process that took place. As will be demonstrated by the analysis, the e-mails contributed to strengthen impressions that initially emerged through our observations of the working group, and which concerned the identification of translation rules that were performed. The e-mail communication then contributed to extend the base on which these interpretations of the material was made.
Our reliance on the instrumental translation theory further implies that the performance takes place within a 'definition of the situation' (Goffman, 1974) in which the contents of the new model is the main subject of attention. Based on our identification of the themes that were addressed in each meeting, however, we were enabled to investigate the orientation of group discussions in more detail. For example, when the discussions in the group were linked to themes as 'AMPP' or 'current practice', we considered them as orientated towards contents. When the discussions were linked to themes as 'group processes' and 'group coordination', we considered them as orientated towards change design. Our observational data indicated that an orientation towards design of the change process was more salient than the orientation towards the contents of the 316 E. A. NILSEN AND A. G. SANDAUNET new model of manpower planning. The impression of an issue that needed further attention was confirmed through the use of 'member reflection' (Tracy, 2010) at the meeting where we were invited to present our 'preliminary thoughts' on the process. This move was deliberate and part of our attempt to validate our observations. In accordance with the directive content analysis (Hsieh & Shannon, 2005), we then conducted a revision of the initial coding framework that was based on the instrumental theory of translation, and extended the framework to include concepts from change design literature.

The Unfolding of the Translation Process
The following analysis demonstrates that the translation process studied can be described in terms of the rules proposed by the instrumental translation theory. However, a switching between which translation rule that gained support and was performed in different parts of the process, was noted. This leads to the impression of a translation process that unfolded in three different phases. The three phases partly followed each other, partly took place at the same time. Each of them represents a chronology of meetings in the group and included a number of sequences.
The analysis further demonstrates that the performance of translation rules took place within the context of both an orientation towards the contents of the new model and towards the design of the change process. Below, we describe the three phases and outline our observations in further detail.

Distancing
We use the label 'distancing' to describe the group's approach to AMPP in the initial meetings. Based on the expected orientation towards the contents of AMPP, we assumed that information about the model, and particularly the new dimension of manpower planning, was introduced and addressed in the initial phase of the work. Examples of such introduction could be resolutions made by the board in the health trust or collected knowledge about AMPP from other health trusts. However, it took time before the group members were presented and 'exposed' to the contents of AMPP, and particularly the activity dimension of the model appeared absent as a topic for discussion in these initial meetings. The initial meetings were first and foremost used to clarify the expectations with which the group members entered the group work, and to go through the mandate. Additionally, we did not register initiatives by the doctors that could be interpreted as a request for more information about AMPP. The doctors expressed worries about the intended change in manpower planning as part of a constant stream of 'cost-cutting' efforts initiated by the top management, but paid less attention to its particular contents. It is on this basis our impression of distancing the new model emerges.
However, not explicitly addressing the contents of the new model does not imply that the performance of translation rules cannot be identified in this part of the process. For example, some departments in the health trust had started to make changes of manpower planning on their own initiative, not as part of the AMPP implementation, but due to experienced challenges related to their manpower planning. One group member worked in one of these departments. Health professionals from other departments that were involved in such initiatives were also invited by the group coordinator to present how they were addressing the experienced challenges. In addition, patterns revealing challenges related to the delivery of services in the health trust as a whole, and assumed to be partly caused by insufficient manpower planning, were presented by one of the representatives from the HR-unit. Exposing the group members for various efforts of improving manpower planning and challenges that could be related to current practice of manpower planning might be interpreted as 'something' (Røvik, 2007) that brings inspiration to start providing contents to AMPP. This explorative approach resonates with the notion of alteration as a way forward in translation processes, since alteration implies multiple degrees of freedom in the modification of an idea (Røvik, 2007;Waeraas & Sataøen, 2014).
From this point of view, the initiation of a translation process can be argued to take place in spite of the impression that the group discussions did not pay direct attention to the contents of the new model. Instead, the alteration described above appears as connected to an orientation towards the design of the change process. During our talks with the group coordinator, she told us that it was her deliberate choice not to introduce AMPP with reference to an 'original version' or how others have translated this model of manpower planning. Her view was that AMPP 'should not be forced upon the doctors'. Both the group coordinator and the group participants then, appeared as more oriented towards how to conduct the process than towards the contents of AMPP. Emphasizing a need for involvement, the group coordinator stated that the doctors' ownership to the change was her main concern. The attention paid to change design is also supported by the fact that we linked several of the discussions in this part of the process to themes as 'group process' and 'other processes in the trust'. The next step of 'rolling out' involvement still appeared as challenging. Initiated by the group coordinator, for example, the group participants went on to prepare an exploration of current practice on manpower planning among doctors in the health trust. Two participants, one doctor and the administrative consultant who had worked with manpower planning in her own department, described an experience of not 'speaking the same language' when they were assigned a task of constructing an interview guide in order to do this exploration. The group then entered an exchange of views through which they concluded that it would be difficult to complete the exploration if they were not able to develop a common reference frame that they all could agree upon. After the meeting, the need for common reference frames was commented in the minutes that the group coordinator distributed, indicating that it was a challenge that had to be addressed. Despite this agreement, the challenge was not presented as a task that the group was requested to engage into in the upcoming meetings. It leads to the impression that openings for such discussions were not provided. As such, periodically taking a step back (Raelin, 2001), which we have previously described as an important element of an involvement-oriented process, did not reveal itself. Our example illustrates that apparently, the participants in the group were 'placed' in an 'involvement-friendly' process, but without managing to activate tools that could indicate its realization.
At the same time, the toolkit of the project methodology was rolled out: the mandate was introduced and presented as a request from the top management. Already at the first meeting the date of the deadline for delivery of a final report with recommendations to the management team, was decided. The general tone was that 'we have a lot to do, and 318 E. A. NILSEN AND A. G. SANDAUNET not much time'. A meeting schedule was set, and minutes from the meetings were distributed to all the members of the group. The most common statement among the participating doctors was that they had not had time to look at it. As a further illustration of the influence of project methodology, the group coordinator informed at the fourth meeting that the deadline had been speeded up by the steering committee, and that the report was due to delivery one month earlier than initially decided. As such, the explorative character of alteration had to take place within an emerging presence of project methodology and attention paid to project progression.

Treatment
We describe the next phase as 'treatment'. In this phase, the group started to explore current practices of manpower planning through their own investigations, and to discuss the contents of the new model. Invited presentations also went on in group meetings, and there was input on 'current practices' until the tenth meeting. In the wake of their own investigations and continued inputs from other departments, the group entered discussions through which we observed that potential improvements of manpower planning were commented on and made to an issue by the group participants. Several members of the group also referred to manpower planning as based on tradition and being partly conducted on 'autopilot', pointing at a lack of overview and masterplan of planning as the main reasons for this pattern. The doctors themselves described how the current plans did not account for scheduled absence, for example when specialty registrars were absent due to scheduled and compulsory courses, and at the same time were written into the roster. Comments that confirmed this 'status quo' of manpower planning were following such statements. In total, an increased awareness of the shortcomings in existing practices emerged in these examples of local knowledge on manpower planning. We consider this awareness as a step towards treatment. We still traced two different views on what should be the contents of the new model. Efforts made by the representatives from the HR-department in order to demonstrate the potential of using statistics in the pediatric and X-ray departments respectively, indicated a view on standardization of tasks as the 'road' to improved manpower planning. In these utterances a performance of copying can be traced, succeeding the impression of alteration proposed as characteristic in the distancing phase. Copying involves the transfer of a construct as accurately as possible, with no or few changes from the original concept (Røvik, 2007;Waeraas & Sataøen, 2014). The emphasis placed on statistics by the representative for the HR department is particularly decisive for our impression of copying. The emphasis links to the calls in the resolutions from the board in which the need for another methodology that limits individual choices in treatment decisions, was emphasized.
Such suggestions prompted counter arguments from the doctors. The doctors' response was to address the context-dependent character of patient work and to emphasize that statistics and numbers had to be used with caution. They emphasized that each consultation is unique, and that the gain of standardizing treatment was limited. According to them, every doctor must use his or her experience with a variety of conditions in order to plan future activity. At meeting eleven one also expressed that: 'On many occasions experience is as important as statistics and figures'. The doctors thereby proposed an alternative approach to improved coordination, which is based on their professional assessment of each incident. This alternative approach to manpower planning deviates from the performance of copying and 'keep up' the impression of the process also displaying alteration.
The group coordinator had supported the doctors in previous sequences when they repeatedly argued that numbers must be used with caution. On this occasion, however, she responded that 'experiences are often related to feelings, and does not necessarily represent a neutral dimension'. As such, the response that was uttered represented a rejection to the effort to develop an alternative approach of how expected activity can be scheduled. The sequence then closed, due to the time schedule. It illustrates clearly how the administrative staff and the doctors deviated in their views on use of numbers and statistics as a basis for scheduling future activity. We are left with the impression that copying and alteration challenged each other as alternative versions of the new model. Independent of the 'competition' between copying and alteration, the performance of these rules appeared to take place in the context of an orientation towards the contents of the new model, and which was based in an agreement among the group participants that they had to change and improve their manpower planning. However, in the next meeting (number twelve of sixteen meetings), the group directed attention to the report and thereby the delivery provided by the working group. This turn of the group attention took place at the same time as both the group coordinator and the group participants had agreed that the discussion at the last meeting was not concluded and needed to be continued. Similar to what we commented in our outline of the distancing phase, agreements on further discussion represent occurrences that are not 'picked up' in the group process. We consider this observation as a further expression of the group not taking a step back in order to 'ponder the meaning' behind the different views that are becoming more visible. In other words, involvement does not envision itself in concrete actions and underscores our impression of the 'lack of' practical tools for reflection, at the same time as the attention also turns to the report. Project progression thereby emerges as the addressed matter.

Simplification
We describe the last phase we have identified as 'simplification'. The description refers to what AMPP 'became' in the group and is based on the group members' work with the report, and the report itself. We consider the report as the most important source for demonstrating what AMPP turned into during the process in the working group. It is also the important link between this initial work of introducing AMPP to the doctors and the further steps that were to be taken in the health trust in order to implement AMPP. Using the term 'simplification', we direct attention to what we consider a gap between the contents of the discussions in the group, particularly in the treatment phase, and the contents provided to AMPP in the report.
The summary of the report describes the report as a 'suggestion of a planning process based on defined activities', and consequently, a version of AMPP that can be implemented throughout the organization. The report outlines three processes or 'steps' for the work with elaborating and operating activity-based manpower plans 320 E. A. NILSEN AND A. G. SANDAUNET across the departments. Our argument for a simplification in the 'report version' of AMPP relies particularly on how AMPP is suggested as a point by point listing, almost a recipe, of the tasks that are included in elaborating an activity plan (for example, which activities to be conducted in a department or unit), medical competence, and manpower plan. Statistics and historical figures are considered as a base for the estimations. Expected timeconsumption of pre-determined activities shall also be specified. Finally, the need for doctor resources emerges as a result of the prior made calculations. This stands in contrast to the different views demonstrated when the group addressed activity as a basic component in future manpower planning. For example, what is not described in the report, is that several of these 'listed points' were object of considerable discussion in the working group, prompted by the different views that were made visible in the treatment phase, and particularly related to how activity can be predefined. In other words, when describing the first process, the report does not bring to the fore experiences from the working group that direct attention to potential challenging clarifications of activity. These 'difficult issues' are not made visible in the reportor the report does not tell us that they are there. Based on this observation, the version of AMPP that is presented can be argued to represent a simplified version of what was discussed in the group. In terms of translation, the report version can be described as the result of a copying of what we in an earlier section of the paper presented as the intentions with AMPP provided by the top management. The efforts of alteration argued to be traced in the distancing and the treatment phase are not visible.
We read the report primarily as answering to requests of progression from the steering committee. Parallel to the work taking place in the group, the steering committee requested drafts of the report. Based on our own observations of how the steering committee commented on these drafts, we also knew that the group coordinator was requested to write a report addressing the mandate in a simpler and more focused manner. This included expectations of clear interventions and a plan of action for further implementation. In addition, the group coordinator provided information to the group that they did not have much time to complete the report. Again, we note that the requirements of project progression puts regulations on the group.
At the last meeting, one of the doctors expressed a concern about not being in control of the contents of the report due to the time pressure. Other participants stated that they might not have the time to read the report before submission. However, the response to the steering committee was that the requirements were met, indicating that these signs of resistance were not picked up by the group coordinator. The influence of project methodology is once again illustrated. Table 2 gives a summary of the main findings regarding translation-and change design activities.

Discussion
This article started out by asking how the instrumental translation framework accounts for micro-processes of planned change. An empirical investigation of how a change in manpower planning among doctors in a Norwegian health trust was conducted in order to assess the contribution from this theory. A first finding from the analysis is that the performance of translation rules could be identified, but also that there was a switching between which rule that gained support in different parts of the process. A second finding is related to the context in which the translation rules were performed, indicating an orientation towards design, and demonstrating that project methodology emerged as a key factor of influence. Considerable attention was also paid to involvement in the initial part of the process. However, the application of the bottom-up approach became less influential in the latter part of the group work. These observations lead to the interpretation that project methodology with its emphasis placed on deadlines becomes the influencing factor during this change process. It seems like what Liguori calls the pace-dimension of project methodology (Liguori, 2012), has a stronger bearing than the need to 'step back' in this particular process.
Considering this strength of project methodology as an expression of the influence of a managerial logic, our findings may also be channelled into a discussion of the factors influencing the future contents of resource planning of health personnel. The superiority of the project methodology could indicate that the professional logic did not dominate the translation process in the group studied here. On a general level, such limited influence might have implications for the contents of work schedules that are developed for medical work. As such our finding does not confirm recent suggestions about an assimilation of the managerial logic into a professional logic (Exworthy et al., 2019).
Based on our findings, we move beyond the question whether the performance of translation rules links to more or less presence of translation competence and suggest that it is necessary to extend the scope of potential competences at play. In this context of change, project methodology competence and reflection competence can be argued to interact with translation competence. Since Røvik's (2016) argument only addresses an interrelationship between translation rules and translation competence, The contents of AMPP became the centre of the group's attention

Copying:
The administration's effort to define activity as something to be quantified and standardized.
Alteration: The doctors' arguing that activity is context-dependent and based on experience.
Extensive use of knowledge about existing practice and context Intensified project progression when steering committee decided on a shorter deadline and requested a report with a simple and focused design, followed by a clear plan of action No interventions such as 'stepping back' to reflect

Simplification
Gap between contents of the discussions in the group, and the contents of the report Copying: The description of AMPP in the report was in accordance with the expectations and descriptions referred to as 'the original version' of AMPP The delivery (the report) was according to the requirements and deadline No 'reflection sections' in the report 322 E. A. NILSEN AND A. G. SANDAUNET the next section provides a more in depth interpretation of aspects related to our second finding. Our argument, following the notion of Guiette and Vandenbempt (2017), is that there is a 'taken-for-grantedness' in change implementation, which is not fully accounted for by the instrumental theory of translation.

Taken-for-granted Enactments of Managing Change: Involvement and Project Progression
According to Guiette and Vandenbempt (2017) it makes sense to expect that organizations that are exposed to change initiatives develop taken-for-granted enactments of managing change, put together from different 'recipes' on the subject. Guiette and Vandenbempt call this phenomenon 'change managerialism', which is interesting in a context of continuous change (Maholtra & Hinings, 2015;Meyer & Stensaker, 2006). Our health trust has gone through numerous extensive change processes the last years, in line with-and recognizable throughout the public health service in Norway. Therefore, it is reasonable to assume that the trust would tend to make the same choices each time a new change is initiated, and thereby develops taken-for-granted enactments of change management.
In our case, we consider the taken-for-grantedness as displayed through the orientation towards project progression and involvement, indicating that these change design elements represent certain stakeholder issues that frame the conduct of change processes in the health trust. Regarding project progression we root the taken-for-grantedness in the notion of a 'projectification' of society (Aubry & Lenfle, 2012;Bergman et al., 2013). For example, there has been a rapid expansion of projects in a variety of settings (Sahlin-Andersson & Söderholm, 2002), and studies show that also public sector organizations increasingly initiate projects aiming at changing work, structures, or cultures (Fred, 2015;Löfgren & Poulsen, 2013;Westrup, 2013).
When it comes to involvement, we root the taken-for-grantedness in fundamental thoughts on participation and involvement in the Scandinavian model (Børve & Kvande, 2012;Gustavsen, 2007). For example, the current emphasis on empowerment in efforts to develop organizational capacity (Argyris, 2001) can be argued to reflect the influence of this bottom-up approach. However, while a stakeholder issue in terms of efficacy might underpin an orientation towards project progression, the drive towards involvement and participation might rather be connected to a need for legitimacy that is considered to be particularly present in public organizations (Fernández-Alles & Llamas-Sánchez, 2008). Furthermore, we suggest that poject progression has become an integrated part of a 'project-methodology-recipe' that is filled with tools to create action points, whereas the ideal of involvement finds itself less materialized into certain reflection techniques ready to be 'rolled out'.
While both project progression and involvement were emphasised as important areas of attention for the group that was studied, we were not able to register the same awareness of how the contents of AMPP ought to be handled. Therefore, and despite the fact that translation of managment ideas have been given increased attention the last few years (Sturdy et al., 2019), we do not consider translation competence as part of takenfor-granted enactments of managing change in our specific context. The overall impression is instead that the relative strength between 'possible' taken-for-granted enactments of managing change is unevenly distributed. Consequently, project methodology is present both rhetorically and practically. Reflection is rhetorically present, whereas translation issues are not rhetorically present, and our observations of practice indicate that translation competence is missing. As such, taken-for-grantedness creates patterns of what we will describe as discontinuities, in the meaning of inconsistent, incomplete or absent use of translation-and reflection competence. Our argument is that this discontinuity is produced through the uneven toolkits related to the taken-for-granted ingredients of change enactments and through a seemingly low awareness of translation competence as such among change agents.
Regarding reflection competence, for instance, the discontinuity in the application is traced in how the talk and intentions related to involvement stand in contrast to the absence of concrete reflection interventions. The group does not ever 'take a step back' to reflect upon their differences. This has similarities with Maholtra and Hinings' (2015) description of awareness-blocking as an aspect of carrying out change, raising the question whether the practice of reflection is sufficiently picked up in change processes. Hence, the use of this competence seems incomplete or inconsistent.
Similarly, we observe several examples of discontinuities in relation to the application of translation competence. One comes to sight in comparing the distancing-and the simplification phase. Here we find that the group ends with a copied version of AMPP in the simplification phase, in spite of indications of alteration in the distancing phase. As stated above, our interpretation is that alteration in the distancing phase is a result of the emphasis placed on involvement, and not of a conscious choice of a translation rule stemming from translation competence. The other discontinuity is observed when comparing the treatment-and the simplification phase. The version of AMPP promoted by the doctors in the treatment phase, is not present in the final report. This is understood as a result of the influence of project methodology competence. Hence, both observations illustrate that even though translation rules are performed, translation competence seems absent as a driving force for the course of events.

Contributions
Based on the analysis and the discussion, a theoretical and a practical contribution are proposed to the literature on change management.

Theoretical Contributions
The theoretical contribution concerns a further development of the instrumental-theoryof-translation-framework, and is developed along two lines of reasoning. First, the switching observed between the performance of different translation rules during the change process directs attention to Røvik's (2016) notion of 'translations modes', which presupposes that the translators of a certain change initiativewhen using translation competencechoose one specific mode accompanied with one specific translation rule, and stick to it during the whole process. However, the switching found in this process opens up for critical remarks and interpretations that can contribute to point a direction for further development of the theory.
According to the framework, the performance of several translation rules during the change process can be interpreted as an expression of a lack of deliberate use of translation competence in the working group. The underlying assumption here is that the use of translation competence in the working group would have led to the performance of only one translation rule throughout the whole process. At the same time as this interpretation finds support in the overall argument in the discussion part of this article, which states a low awareness of translation competence in the working group, the interpretation also contributes to illustrate parts of the internal coherence of the theory.
Another, and somewhat contradicting interpretation, is that the analysis conducted here suggests that the probability of discovering the performance of several translation rules increases the closer the researchers are to the micro processes of change. This comprehension is based on the assumption that the micro processes of a change are too finemeshed with actions to be characterized by only one translation rule within one translation mode. Instead, a more dynamic interplay between the performances of translation rules is revealed, corresponding to findings in other studies (for example Øygarden & Mikkelsen, 2020). As such, this interpretation is not compatible with the outline of 'translation modes' (Røvik, 2016). This further raises the question whether the framework relies on a certain distance between the actual translation work being done, and the level in which the translation analysis is carried out. Rather than pointing to a lack of internal coherence of the theory, this line of reasoning suggests a 'lack of coherence' between theory and the practical conduct of change.
This argument could be taken even further to comprise the relationship between translation rules and translations. Whereas the translation rules provide wide categories for probable versions of the idea being subject to translation, the concept offered by the instrumental theory of translation to get access to the relevant micro level interactions through which ideas are translated, is narrowed down to translation competence. Through Røvik's (2016) outline of scope conditions, we consider translation competence as providing quite a limited scope for the search of translations. As with the relationship between modes and rules outlined above, an internal coherence between the translation rules and how they relate to the scope conditions can be traced, representing what the researcher has to look for in order to spot translations that can add up to the performance of one rule or another. However, the challenge, as we see it, is rather that the concept of translation competence does not seem to fully meet the expectations of the researcher to capture the translation processes as they unfold in practice.
The point is partly revealed through this article's analysis in the sense that an effort was made to use indicators of translation competence. For instance, 'knowledge of the idea to be implemented', was used as an entrance to focus and identify translation work, in order to substantiate that copying and alteration had taken place. However, as the inclusion of change design competences in this study illustrates, we experienced a gap between the tools provided from the concept of translation competence and the activities and interactions we identified as relevant to understand the development of the different versions of AMPP. Again, we provide a line of reasoning that suggests a 'lack of coherence' between theory and the practical conduct of change.
To strengthen this argument we suggest that a variant of this point is being illustrated in Øygarden and Mikkelsen's article (2020). While the rules were used to conceptualize that the contents of an ICT-supported task planning system for hospital physicians changed through translations, the translations themselves were accessed through the concepts of 'editing practices' (Teulier & Rouleau, 2013) and 'readiness for change' (Armenakis & Harris, 2002). Hence, in both examples, the researchers turn to implementation theory in order to access profound identities of some distinguishable quality that can be matched by empirical evidence.
This interpretation of the relationship between modes, rules and translations, on one hand illustrating the internal coherence of the instrumental theory of translation, on the other raising questions concerning a 'lack of coherence' between theory and the practical conduct of change, gives reason to investigate whether the framework meets the request for a micro approach to translation work. This call is in accordance with other calls for research on translation work (Mueller & Whittle, 2011;Teulier & Rouleau, 2013).
The second line of reasoning that concerns a further development of the instrumental theory of translation, relates to the fact that our analysis failed to demonstrate that the performance of translation rules took place mainly in the context of an orientation towards the contents of the new model. This finding raises questions concerning Røvik's (2016) specification of the translation rules' scope conditions. For instance, one scope condition for copying is outlined as follows: 'The more explicit, less complex and less embedded a desired source practice, the more appropriate copying will be as a translation rule in knowledge transfer' (Røvik, 2016, p. 12). Hence, the attention is directed towards the translators' knowledge of the idea. Based on this article's analysis we suggest extending the scope conditions to include change design components. The analysis has demonstrated that taken-for-granted change designs influence translation possibilities, and suggests interaction of different competences in planned change.
A way forward to further develop the theoretical framework could be to elaborate ways of taking into account how taken-for-granted change designs influence translation possibilities. For instance, it is possible to envision that the performance of certain translation rules requires certain designs, and that certain designs may hamper or facilitate the performance of certain translation rules. Translation as alteration implies multiple degrees of freedom in the translation of a new model. According to Røvik alteration becomes the appropriate translation rule 'the more tacit, complex, and embedded a desired source practice' is (Røvik, 2016, p. 14). However, we suggest that an additional scope condition could be that in order to succeed with alteration, a certain level of reflection competence is required. Alteration is a rule that may be used when a new model is seen as a source of inspiration to local innovation and change. From innovation literature (Kesting & Ulhøi, 2010) we know that such processes would benefit from an involving and reflecting methodology. Furthermore, a copying strategy would most likely best be facilitated by a project management methodology because the task here is to reproduce as accurately as possible according to an 'order', and the toolkit of project methodology enables such a top-down approach.
Combining different scope conditions derived from reflection-and progression competence with scope conditions in translation competence, will possibly contribute to bridge the gap between theory and the practical conduct of change, which we previously commented. It will also provide more clues for the researcher to follow in order to get access to the translation work taking place during change. Although the concept of translation rules, in particular, has become a popular analytical framework, this discussion raises the question whether the concept of translation competence will suffer from 326 E. A. NILSEN AND A. G. SANDAUNET marginalization unless the instrumental theory of translation expands and elaborates further on scope conditions for the rules to be performed. However, extending the scope conditions as we are suggesting, is in line with Røvik's own description of defining scope conditions of theoretical constructs as an 'enduring task' (2016, p. 11).
Our study demonstrates the relevance of this point, at the same time as it brings together the two approaches to translation work identified by Cassell and Lee (2017). In their overview of key studies that focus explicitly upon idea translation (2017), the authors divided between those studies using translation rules to investigate the actual idea and the way the idea changes, and those where the focus is upon the processes by which translation happens. We suggest that the processes by which translation happens must be taken into account when trying to understand the translation work 'behind' a set of translation rules. A way of doing this is for instance to integrate change design factors characteristic for the processes by which translation happens, as part of scope conditions for translation rules. In other words, to fully grasp the potential of translation competence, it seems like one has to understand the interrelationship between translation-and change design competences.

Practical Contributions
Our second contribution concerns the practical level, and more precisely the taken -forgrantedness related to competences at play during change implementation. For change agents this is important insight to take into consideration. If the taken-for-grantedness leads to discontinuities, understood as a continual inconsistent use of competences among change agents, in our case in favour of project methodology competence, how will this impact the results of planned change? Even though our study gives us no opportunity to generalize, it may still allow us to pose this question. For example, translation competence directs attention towards the contents of the idea and the knowledge relevant to fully grasp the complexity often inherent in management ideas. One could expect this to be a core aspect of managing change. Instead, it seems like process is more important than contents. This calls for a re-balancing-act. In order to turn the discontinuities to consistencies, we suggest that a future agenda for the change management field should be to raise the awareness of potential discontinuities and their interrelatedness with taken-for-granted change designs. On the practical level, this implies for example such matters as to raise the awareness among change agents of the need for a corresponding reflection-and translation toolkit to that of project methodology, as well as being aware of connections between translation rules and change design. The influence of projectification is, for example, argued not to be unproblematic. The felt time pressure can lead to shortcuts that are solely motivated by a desire to stay on track (van Berkel et al., 2016). Hence, the use of project methodology competence on 'autopilot' at the expense of the two others, may hamper the possibility for disruptive features in change processes, and thereby more radical or transformative approaches to translating the contents of change (Seebode et al., 2012).
As a prolongation of the reasoning above, we would argue that the instrumental translation framework represents a useful contribution in helping change agents by offering a way of building bridges over different logics in organizations. In many respects, our study describes a typical challenge in health-care change initiatives, namely the confrontation between different worldviews or institutional logics (Thornton et al., 2012), with the doctors representing one and the administrative staff representing another. As one of our informants said 'we do not speak the same language'. While project methodology competence, being associated with a managerial logic, seems to sustain the cleft between these different professional groups during change, we see a potential in the concept of translation competence in bringing different logics to closer collaboration through focusing on contents and knowledge of the idea to be implemented. Bluntly speaking, a focus on the contents of the idea may have a greater potential to 'bring the two languages closer together' than what can be expected from a focus on project progression. The chance is, that by raising the awareness of translation competence, the repertoire of ways of working to obtain change will not only be extended, it will also be more aligned with a general expectation of knowledge creation as part of change.

Conclusion
This study addressed the usability of the instrumental theory of translation. More precisely, it represents an effort to combine the concepts of translation rules and translation competence in an empirical analysis, and to discuss how the framework accounts for micro-level translation work. Although the participative research approach and the context bounded character of the data provided have limitations and must be closely considered, important issues for further investigation can be raised. On the theoretical level, we have identified coherence between the instrumental translation framework and the practical conduct of change as one such issue. Another is the specification of translation rules' scope conditions. On the practical level, we call for more awareness towards the impact of taken-for-granted enactments when managing change.